Galaxy major mergers (mass ratio >4:1) are thought to play an important role in galaxy evolution. Galaxy close pairs and hallmark tidal features are frequently-used empirical merger identifiers and are used to quantify merger rates. The observability timescale (Tobs) of these empirical signatures is likely the key (and presently most uncertain) variable for deriving merger rates. To improve our understanding of these observability time scales, we use mock WFC3-F160W images with realistic noise of 4800 mergers among massive galaxies (log Mstellar/Msun >10) saved at ~10-20 Myr time resolution and spanning redshifts 1<z<3 from the Horizon-AGN hydrodynamic simulation. For each merger in bins of redshift (dz=0.5) and stellar-mass of the host galaxy (dLogM = 0.5), we compute the projected distance (Rproj) between the host and companion galaxies at each timestep and quantify the cumulative time observable in a window 5 kpc < Rproj < 30 kpc as a close pair. Our preliminary results suggest that the median Tobs,pair decreases from ~200 Myr to ~100 Myr during 1 < z < 3, broadly consistent with an evolving timescale [~(1 + z)-2] argument proposed by previous simulation-based works. However, we notice a factor of two variance in Tobs values at a fixed redshift bin, which we suspect may be due to different intrinsic merger orbital configurations. Simultaneously to quantify the Tobs,tidal, we developed an automated routine to model the light profile of sources within each timestep of the mergers with a single-Sersic model using GALFIT. We process the model-subtracted residual images through a feature extraction routine to extract flux and area wise contiguous pixel regions and quantify the surface brightness and area of the resultant tidal features. We plan to quantify the Tobs,tidal as a function of key variables such as redshift, stellar-mass of host galaxy, and mass ratio to better understand the link between the observed tidal signatures and the properties of progenitors causing them.